Skip to main content

Slowing productivity? AI might need to take the wheel.

AI has the potential to support stronger labour productivity in Australia, while simultaneously transforming occupations in the years to come.

The latest Deloitte Access Economics Employment Forecasts report highlights the productivity challenge currently facing the Australian labour market. 

With nearly 400,000 Australians finding jobs over the past 12 months, the strength of the Australian labour market is an achievement that should be celebrated. However, since its peak in March 2022, Australia’s labour productivity has fallen by 5.7% and productivity in the non-market sector is at a near 20-year low.

AI (including automation, GenAI and Agentic AI) may hold the key to our productivity dilemma. It is, however, important to note the impact of AI is unlikely to be uniform across the economy, with occupations and underlying tasks likely to be transformed differently. Some of this may be pure efficiency savings, though in many cases the time freed up will allow a higher quality of work, or new kinds of work to be performed.

To systematically capture the variety of outcomes, the latest occupation projections in Employment Forecasts integrate Deloitte’s Work Analyser tool. Sarah Rogers, Deloitte Human Capital partner, notes that “Work Analyser helps our Workforce Strategy experts rapidly assess potential AI impacts on jobs, tasks, skills and workforce size, shape and cost – identifying and quantifying efficiency opportunities. This insight helps organisations to make strategic and integrated choices about workforce and technology transformation, and where work needs to change.” 

The AI automation score from Work Analyser informs the magnitude of potential AI impact on the workforce, while the human necessity score informs the degree to which it is necessary that a human performs the work. While scores often have an inverse relationship, this is not always the case and can inform decisions about AI and reskilling strategies, including whether a role should be fully automated, augmented, or assisted by AI. 

While many occupation groups are dispersed across the spectrum of human necessity and AI automation, there is some clustering, reflecting similar underlying tasks and skills. In particular, AI is expected to have a more significant impact for ‘clerical and administrative workers’. This occupation grouping tends to have relatively lower needs for human involvement and relatively higher AI automation potential. 

Within this occupation grouping, ‘call or contact centre workers’ has one of the highest AI impact scores, with significant capacity release opportunities. Agentic AI has the power to automate routine inquiries and processing, with AI-enabled assistants collaborating with humans to analyse data and accelerate solutions. As a result, employment in this occupation is forecast to decrease by 1.1% per annum in the decade ahead. 

‘Professionals’ is a more diverse occupation grouping including business professionals, scientists, educators, health care workers, and engineers. The role of AI for ‘professionals’ is enormous, but its impact differs from that of ‘clerical and administrative workers’. Instead of reducing the demand for an occupation, AI is likely to provide opportunities for automating routine tasks and freeing up time for higher-value tasks, including new types of work.

For instance, the work of ‘chemical and materials engineers’ may be redefined by advanced AI systems to focus more on design, simulation and development of high performance materials. Thus, AI developments are likely complementary to growth in this occupation as potential efficiency gains will likely re-shape the day-to-day tasks of the role. This represents the renewed potential for change in occupations that can provide innovation in the economy. 

AI has the potential to support stronger labour productivity in Australia, while simultaneously transforming occupations in the years to come. Looking ahead, as the Work Analyser tool yields deeper insights, future editions of Employment Forecasts will continue to integrate Work Analyser’s outputs into its occupation projections to systematically model the impacts of AI technologies on Australia’s workforce.

This newsletter was distributed on 28th May 2025. For any questions/comments on this week's newsletter, please contact our authors:

This blog was co-authored by David Li, Graduate Economist at Deloitte Access Economics

Click on the links below to read our previous Weekly Economic Briefings: